Prevalence of pelvic floor dysfunction in recreational athletes: a cross-sectional survey

Introduction and hypothesis Pelvic floor dysfunction (PFD) affects many women and participation in elite sport and high-impact exercise has been reported as a potential risk. However, few studies have investigated the effects of exercising at recreational levels on PFD. Our aim was to investigate levels of PFD in women exercising at, or above, UK guidelines for health and compare them with levels in non-exercisers. Method Data on levels of PFD and potential risk factors (age, hormonal status, body mass index, constipation, parity, forceps delivery, and recreational exercise) were collected using a cross-sectional survey distributed via social media. The International Consultation Incontinence Questionnaire (ICIQ) Urinary Incontinence Short Form was used to estimate prevalence of urinary incontinence (UI). Selected questions from the ICIQ vaginal symptom and bowel symptom questionnaires were used to estimate prevalence of anal incontinence (AI) and pelvic organ prolapse (POP). Logistic regression analysis was used to compare exercisers and non-exercisers after adjusting for potential confounders. Results We recruited 1,598 adult women (1,141 exercisers and 457 non-exercisers). The majority were parous. High prevalence of UI (70%), AI (52%) and POP (18%) was reported. No significant association was found between recreational exercise and PFD despite adjustment for confounders, or further investigation regarding exercise involving impact, although some increased reporting of AI was seen in those exercising for over 10 hours per week. Conclusion High levels of all PFD were reported but no significant association was found between recreational exercise and symptoms. However, data suggest that women modify their exercise regimes as required. Few symptomatic women sought professional help. Supplementary information The online version contains supplementary material available at 10.1007/s00192-023-05548-8


Introduction
Pelvic floor dysfunction (PFD), which includes urinary incontinence, anal incontinence and pelvic organ prolapse (POP) [1], causes embarrassment and distress, limits many aspects of life [2] and affects many women [3].It is accepted that childbirth, obesity and aging are risk factors for PFD [4] but recent evidence suggests that the prevalence of urinary incontinence in young, nulliparous athletic women is 2.77 times higher than in their sedentary counterparts [5].Other reports suggest that high-impact activities, e.g.cheer leading, may be linked with increased levels of anal incontinence (AI) [6].However, regular participation in sport and exercise confers multiple health benefits [7,8].Current UK recommendations are that adults should exercise at moderate levels or above for a minimum of 150 min each week, over three sessions [9].Urinary incontinence (UI) can be a barrier to exercise [10] and concern regarding potential risks to the pelvic floor, as reported in elite athletes, may cause health professionals and women to question the safety of engaging in sport and exercise, for fear of aggravating symptoms or increasing the risk of developing PFD.Although studies have investigated whether the risk of PFD is higher in elite athletes than in sedentary individuals [11] and in younger women [6,12], only a few have reported levels within a broad range of recreational athletes [13,14].Therefore, the objectives of this study were:

Participants and recruitment
Adult women were invited to take part via advertisements, which were widely distributed on social media networks (Facebook, Twitter, Instagram and LinkedIn) and using snowball methodology (asking people to share the information with others).Posters were also distributed to sports clubs, workplaces and physiotherapy clinics for display on websites and notice boards.QR codes linked directly to the survey.Advertisements highlighted that ALL women were invited to take part: both those who did and those who did not exercise and both those with and those without any pelvic floor symptoms.Data collection took place between 6 May and 31 July 2022.

Outcome measures
To determine the prevalence of UI, AI and POP (as defined by the International Urogynaecological Association (IUGA)/International Continence Society (ICS) joint report) [1], we collected data using patient-reported outcome measures.We used the International Consultation Incontinence Questionnaire (ICIQ) Urinary Incontinence Short Form (ICIQ-UI-SF) [17] in its entirety, and specific questions of interest from ICIQ bowel symptom [18] and ISIQ vaginal symptom questionnaires [19].Inclusion of all three questionnaires in full would have resulted in a prohibitively time-consuming survey, likely to deter participation.
Those who reported "never" in response to the question "How often do you leak urine?" were classified as continent of urine, and severity was defined by the ICIQ-UI-SF severity score [20].UI was further subdivided into stress UI (SUI), urgency UI (UUI) and mixed UI (MUI) based on the answers to "When does urine leak?" Anal continence was identified in those who answered, "always" to the question "Are you able to control leakage of stool or flatus (wind) from your back passage?"Responding "never" to the question "Are you aware of a lump or bulge coming down in your vagina?" was taken to indicate the absence of POP.Awareness of a lump or bulge in the vagina has been associated with the presence of a grade 2 POP, although this may underestimate the true prevalence of this dysfunction [21].Age, menopausal status, body mass index (BMI), constipation (defined as regularly having to strain to open bowels), parity and type of delivery were considered potential risk factors for PFD and possible confounders.
Recreational athletes were defined to be those who met and exceeded the UK Chief Medical Officer's guidelines for healthy living of 150 min a week [9].This was further subdivided into high-impact (sports involving both feet leaving the ground at the same time, e.g., running, high impact gym or trampolining) and low-impact (one foot always in contact with the ground or body weight supported, e.g.walking, cycling, kayaking) or both.
Additionally, participants were asked if they had sought professional help for PFD, and if they regularly performed pelvic floor exercises.The final open question gave participants an opportunity to record comments or additional information regarding previous answers.

Statistical analysis
Statistical analysis was performed in SPSS statistical software version 28 (IBM, Chicago, USA).Demographic data were reported using frequencies with percentages or means with standard deviations (SD).Prevalence was reported as frequency and percentage, and Pearson's Chi-squared test was used to investigate any differences in prevalence of PFD between non-exercisers and exercisers.Missing data were reported.Risk factors for PFD were estimated by logistic binomial regression analysis, reported as adjusted odds ratios (OR) with 95% confidence intervals (CI), with the significance level set to 0.05.

Results
Visits to the survey site were recorded to be 4,985: 3,185 exiting after the first page (survey information).A few individuals then exited from subsequent pages but most who progressed to consent then completed the survey (Fig. 1).Individual IP addresses were not collected due to the anonymising process so it is not possible to calculate participation rate; each visit recorded could represent duplicate visits by the same participant or unique visits.
In total, 1,600 participants consented to take part and submitted data.Two were excluded: one self-identified as a man, noting that they were male at birth, but wished to underline the need for a similar survey for men; and one did not provide key data regarding birth history, menopausal status and exercise history.Data submitted by the remaining 1,598 participants was analysed: of these 1,141 (71%) reported exercise levels above UK guidelines of more than 150 min per week and 457 (29%) did not exercise or were below this level.Most exercisers (921, 81%) reported doing so for over 5 years and 1,041 (91%) exercised more than three times per week, in line with guidelines.Owing to an initial system issue, 8 participants were able to bypass some questions regarding bowel and vaginal symptoms, which is noted within the results tables.

Associations
There were no significant between group differences regarding exercisers and non-exercisers in levels of UI (p=0.352),AI (p=0.182) or POP (p=0.152).Exercisers were less likely to report constipation: 17% compared with 22% of non-exercisers (p=0.019;Table 2) .
After regression analysis using logistic binomial regression to account for other risk factors (age, reduced oestrogen, BMI, constipation, parity and forceps delivery) with non-exercisers (<2.5 h/week) as the reference group, no significant association was found between recreational exercise and PFD.
Risk factors associated with UI included aging, BMI, constipation and parity.AI was associated with age, constipation and forceps delivery.POP was associated with hormonal status, constipation and increasing parity (Table 3).
Further investigation to account for potential effects of exercise involving impact only as opposed to non-impact sport revealed no significant differences in levels of PFD (Table 5).

Pelvic floor exercises and treatments
Pelvic floor exercises were performed regularly by only 646 (40%) participants.
Of those reporting any pelvic floor dysfunction, only 450 out of 1,319 (34%) had sought professional help.Those with symptoms were no more or less likely to exercise their pelvic floors.

Responses to open question
In the final section, 537 participants made comments.These are reported in detail elsewhere [22] but we report, in brief, key illustrative quotations regarding impact of symptoms on access to sport and treatments to manage symptoms.

Pelvic floor symptoms impacting access to sport
Often participants commented that their pelvic floor symptoms were the reason why they could no longer take part in sport and exercise: "I would love to exercise to lose weight, but it is impossible with these bladder issues … it's so frustrating" "…my exercise intensity and frequency have changed since having children due to leakage/prolapse symptoms.Before kids my exercise intensity was high and 6 days/week.Now, I don't engage in high intensity exercise anymore…." Many commented on the negative effects of this, ranging from ensuring that their bladder had been emptied before leaving the house: "I feel like I should always empty my bladder before I leave the house, gym, work etc. to avoid a panic when I need to urinate." to great distress: "It's impacted my life; I can't run anymore ….No one gives a damn because it's only women."

Treatment
Some commented on treatments they had attempted to seek or been offered, and many suggested that pelvic floor muscle exercises did help: "Doing daily regular sustained pelvic floor exercises has greatly improved my symptoms" whereas others found that taking part in sport had helped: "I started to include weight training …. and feel that has helped my pelvic floor enormously." Fig. 1 Flow chart to illustrate site visits and points of exit.NB: as the process was anonymous no unique IP addresses were saved so it was not possible to differentiate exits by unique visitors and repeat visits

Discussion
The objectives of this survey were to determine the levels of PFD in women who exercised at or above recommended guidelines for healthy living and in those who did not and to identify any correlation between exercising recreationally and the incidence of PFD, as previously noted in literature regarding elite athletes [11,12,23,24].All levels of PFD reported were high compared with other studies: UI was reported by 40% of women in a 2015 UK survey [16] compared with 70% of our participants, and AI was reported by only 14% of women in a US epidemiological survey [25] compared with 48% in this survey.However, a recent study investigating the long-term effects of sphincter injuries at birth on AI reported 60% prevalence of AI in the control group (those without sphincter injury) [26] and this level was similar to that found in a group of young, nulliparous women [12].Levels of POP again appear to be greater than that reported in the US epidemiological study [25] but in another recent internet-based survey, 14% of participants reported POP [13] compared with 18% of our respondents.It is likely that there will be some selection bias in an internet survey as women with an interest are more likely to take part, despite advertisements aiming to recruit ALL women.However, it may be that as pelvic health symptoms are increasingly being discussed more openly in the media women are becoming more confident to share information regarding these symptoms.
We found no significant associations between taking part in recreational sport and exercise and PFD other than a small increase in the number of women reporting AI when exercising for more than 10 h/week.However, this should be interpreted with caution given the low numbers of women exercising at higher levels in this survey.It is important to note that many previous studies reporting increased levels of PFD in athletes have investigated the elite population [11], whereas other investigations that have noted significantly higher levels of UI included only young nulliparous women, without the increased extra risks associated with parity and/or assisted delivery in their sedentary cohort [12].Athletes in the latter study reported training on average for 19 h/week, whereas the majority of our exercisers were exercising for less than 10 h/week and a positive association has been reported between volume of physical activity and the frequency of UI [27].However, a previous study also found no significant correlation between UI or POP and exercise, and the only significant correlation reported was between AI and sport [28].
The demands of elite level competition dictate that reducing training levels or modifying load is rarely an option unless there is illness or injury.It is therefore likely that elite athletes, many of whom have never mentioned their symptoms to anyone [29], would not alter their sport or training levels as a result of PFD.In the case of the recreational athletes in our survey, however, comments suggested that women often modified their sports to include lower impact activities or reduced the level of exercise they took part in altogether.This, combined with the lower volume of exercise performed by most of our exercisers, may explain the differences in the results.However, it should also be noted that as the majority of exercisers in our survey have been doing so for over 5 years there is little in this study to suggest that recreational sport at these levels is a specific risk to the pelvic floor.
Finally, the majority of those who reported PFD here had not sought professional help, despite comments suggesting that PFD caused distress.This is recognised and has previously been reported in other studies on both athletes [29] and the general population [16].A major strength of internet-based surveys is the ability to recruit large numbers of participants from a spread of geographical locations.Moreover, although self-reporting of symptoms may be less accurate than using objective measures such as pad tests to detect incontinence or vaginal examination to diagnose POP, this is mitigated by the use of validated questionnaires to predict symptoms.
There are, however, associated limitations, not least of which is the possibility of selection bias, as those affected by the criteria being investigated are most likely to take part, which may increase prevalence levels.In addition, although advertisements asked ALL women to participate, the only inclusion criteria were to be adult and female; this could mean that some were pregnant or possessed disabilities that could have an impact on their pelvic floor function.Further, although we aimed to recruit a diverse population, the majority of participants were educated to degree level or above.It is then even more surprising that most symptomatic participants had never sought professional help.

Conclusion
Overall levels of PFD within this survey are high but there was no association between recreational exercise and the rates of PFD reported.Further longitudinal studies may help to investigate any long-term risks of recreational exercise to pelvic health.However, based on the results of this survey and the multiple health benefits associated with taking part in regular sport and exercise, women and health professionals should be cautious when extrapolating the risks to the pelvic floor associated with elite sport to recreational exercisers.

Page 1 : 6 Exit site n = 2
Study informaƟon n 4985 Page 2: Consent n = 1800 Page 3: Background InformaƟon n = 1756 Page 6: Urinary Symptoms Page 7: Bowel Symptoms Page 4: Birth History Page 8: Prolapse Symptoms Page 5: Exercise Levels Page 11: Useful informaƟon and help sites n Exit site n = 13 No iniƟal consent n = 1 Custom route if nulliparous

Table 1
Characteristics of participantsData presented as frequency (n) and % within group; BMI as mean kg/m 2 (standard deviation, minimum to maximum)

Table 3
Adjusted odds ratios for the relationship between pelvic floor symptoms and risk factors Risk factors: age (years) (18-36 years group as reference); hormonal effects (premenopausal group as reference); BMI; constipation (no straining to defaecate as reference); parity (nulliparous as reference); forceps delivery (non-forceps delivery as reference), exercise group (exercise levels <2.5 h/week as reference group) BMI body mass index, OR odds ratio, CI confidence interval, UI urinary incontinence, POP pelvic organ prolapse, AI anal incontinence *Reached significance in regression analysis.Adjusted for all risk factors noted

Table 4
Adjusted odds ratios to indicate any relationship between levels of exercise and pelvic floor symptoms Non-exercisers as the reference group and adjusted for all risk factors as in Table3OR odds ratio, CI confidence interval, UI urinary incontinence, POP pelvic organ prolapse, AI anal incontinence

Table 5
Adjusted odds ratios for the relationship between types of exercise and pelvic floor symptoms Low-impact group as reference group for relationships OR odds ratio, CI confidence interval, UI urinary incontinence, POP pelvic organ prolapse, AI anal incontinence